[Bug tree-optimization/107809] New: gcc.dg/vect/vect-recurr-1.c etc. FAIL

[Bug tree-optimization/107809] New: gcc.dg/vect/vect-recurr-1.c etc. FAIL

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107809

            Bug ID: 107809
           Summary: gcc.dg/vect/vect-recurr-1.c etc. FAIL
           Product: gcc
           Version: 13.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: ro at gcc dot gnu.org
  Target Milestone: ---
            Target: sparc*-sun-solaris2.11

Since their introduction, the vect-recurr-?.c tests FAIL on 32 and 64-bit
SPARC:

+FAIL: gcc.dg/vect/vect-recurr-1.c -flto -ffat-lto-objects  scan-tree-dump vect
"vectorized 1 loops in function"
+FAIL: gcc.dg/vect/vect-recurr-1.c scan-tree-dump vect "vectorized 1 loops in
function"
+FAIL: gcc.dg/vect/vect-recurr-2.c -flto -ffat-lto-objects  scan-tree-dump vect
"vectorized 1 loops in function"
+FAIL: gcc.dg/vect/vect-recurr-2.c scan-tree-dump vect "vectorized 1 loops in
function"
+FAIL: gcc.dg/vect/vect-recurr-3.c -flto -ffat-lto-objects  scan-tree-dump vect
"vectorized 1 loops in function"
+FAIL: gcc.dg/vect/vect-recurr-3.c scan-tree-dump vect "vectorized 1 loops in
function"
+FAIL: gcc.dg/vect/vect-recurr-4.c -flto -ffat-lto-objects  scan-tree-dump vect
"vectorized 1 loops in function"
+FAIL: gcc.dg/vect/vect-recurr-4.c scan-tree-dump vect "vectorized 1 loops in
function"
+FAIL: gcc.dg/vect/vect-recurr-5.c -flto -ffat-lto-objects  scan-tree-dump vect
"vectorized 1 loops in function"
+FAIL: gcc.dg/vect/vect-recurr-5.c scan-tree-dump vect "vectorized 1 loops in
function"

Dump attached.

[Bug tree-optimization/107809] gcc.dg/vect/vect-recurr-1.c etc. FAIL

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107809

--- Comment #1 from Rainer Orth <ro at gcc dot gnu.org> ---
Created attachment 53944
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=53944&action=edit
32-bit sparc-sun-solaris2.11 vect-recurr-1.c.172t.vect

[Bug tree-optimization/107809] gcc.dg/vect/vect-recurr-1.c etc. FAIL

[ro at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107809

Rainer Orth <ro at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
   Target Milestone|---                         |13.0

[Bug testsuite/107809] gcc.dg/vect/vect-recurr-1.c etc. FAIL

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107809

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
          Component|tree-optimization           |testsuite
             Status|UNCONFIRMED                 |ASSIGNED
   Last reconfirmed|                            |2022-11-23
           Assignee|unassigned at gcc dot gnu.org      |rguenth at gcc dot gnu.org
     Ever confirmed|0                           |1
           Keywords|                            |testsuite-fail

--- Comment #2 from Richard Biener <rguenth at gcc dot gnu.org> ---
Mine then.

[Bug testsuite/107809] gcc.dg/vect/vect-recurr-1.c etc. FAIL

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107809

--- Comment #3 from Richard Biener <rguenth at gcc dot gnu.org> ---
It looks like SPARC should support vec_perm with VIS2, but it's not clear
whether that's enabled.  The testcases definitely lack an effective target
check for the required permute which is a whole vector shift by 1 element
plus the insert of the topmost lane into the first or a two-vector shift
by 1 element.

I've checked and

/* { dg-require-effective-target vect_perm } */

should disable the test for sparc but it's not exactly spot-on I think
(and as said, it looks like -mvis2 should suppport the test)

[Bug testsuite/107809] gcc.dg/vect/vect-recurr-1.c etc. FAIL

[cvs-commit at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107809

--- Comment #4 from CVS Commits <cvs-commit at gcc dot gnu.org> ---
The master branch has been updated by Richard Biener <rguenth@gcc.gnu.org>:

https://gcc.gnu.org/g:d4a320f1ee4c76f09f95bdc583b90a9b049bbba0

commit r13-4846-gd4a320f1ee4c76f09f95bdc583b90a9b049bbba0
Author: Richard Biener <rguenther@suse.de>
Date:   Thu Dec 22 14:20:40 2022 +0100

    testsuite/107809 - fix vect-recurr testcases

    This adds a missing effective target check for the permute
    recurrence vectorization requires.

            PR testsuite/107809
            * gcc.dg/vect/vect-recurr-1.c: Require vect_perm.
            * gcc.dg/vect/vect-recurr-2.c: Likewise.
            * gcc.dg/vect/vect-recurr-3.c: Likewise.
            * gcc.dg/vect/vect-recurr-4.c: Likewise.
            * gcc.dg/vect/vect-recurr-5.c: Likewise.
            * gcc.dg/vect/vect-recurr-6.c: Likewise.

[Bug testsuite/107809] gcc.dg/vect/vect-recurr-1.c etc. FAIL

[rguenth at gcc dot gnu.org]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=107809

Richard Biener <rguenth at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|ASSIGNED                    |RESOLVED
         Resolution|---                         |FIXED

--- Comment #5 from Richard Biener <rguenth at gcc dot gnu.org> ---
Should be fixed now.